Vehicle collisions are often caused when a driver can not see or is unaware of an oncoming object. For example, a tree may obstruct a drivers view of oncoming traffic at an intersection. The driver has to enter the intersection with no knowledge whether another vehicle may be entering the same intersection. After entering the intersection, it is often too late for the driver to avoid an oncoming car that has failed to properly yield.
There are other situations where a vehicle is at risk of a collision. For example, a pileup may occur on a busy freeway. A vehicle traveling at 60 miles per hour, or faster, may come upon the pileup with only have a few seconds to react. These few seconds are often too short an amount of time to avoid crashing into the other vehicles. Because the driver is suddenly forced to slam on the brakes, other vehicles in back of the driver""s vehicle may possibly crash into the rear end of the driver""s vehicle.
It is sometimes difficult to see curves in roads. For example, at night or in rainy, snowy or foggy weather it can be difficult to see when a road curves to the left of right. The driver may then focus on the lines in the road or on the lights of a car traveling up ahead. These driving practices are dangerous, since sudden turns, or other obstructions in the road, may not be seen by the driver.
The present invention addresses this and other problems associated with the prior art.
Sensor data is generated for areas around a vehicle. Any objects detected in the sensor data are identified and a kinematic state for the object determined. The kinematic states for the detected objects are compared with the kinematic state of the vehicle. If it is likely that a collision will occur between the detected objects and the local vehicle, a warning is automatically generated to notify the vehicle operator of the impending collision. The sensor data and kinematic state of the vehicle can be transmitted to other vehicles so that the other vehicles are also notified of possible collision conditions.